Predicting Failures of Banking and Financial Institutions in Nepal: A Comparative Methodological Analysis

Abstract

Identification of the companies prone to various risks is important to regulators, investors, analysts and practitioners. It is imperative to identify early warning indications of impending financial problems and the potential subsequent failure of these companies. This study focuses on failures of Banking and Financial Institutions (BFIs) in Nepal. The main objective of this study is to examine the predictive power of CAMEL ratios to develop a Multivariate Discriminant Model (MDA) and Logistic Regression Model (LRA) in the context of Nepalese BFIs. To achieve this objective a descriptive and analytical research model was designed. The research applies secondary data to predict the predictive failure of the Nepalese BFIs with the support of MDA and LRA. The ratios of non-failed BFIs were relatively stable, contrary to those of failed BFIs under the study. The results indicate that, the overall accuracy in predicting bank failure using LRA is higher than with MDA. Therefore, LRA can be considered to be relatively more accurate than MDA. Finally, the relative contribution of return on assets and non-performing loans to total loans has been found to be a more reliable and predictable ratio with regard to banking failures for both MDA and LRA.